Sports-related concussion is a growing health and socio-economical problem and yet, its biomechanical mechanisms are still poorly understood. Accurate quantification of how mechanical energy from an external impact is transferred into brain mechanical responses is critical for understanding the mechanisms of traumatic brain injury (TBI). We propose to correlate white matter (WM) fiber strain and strain rate accumulated from repetitive head impacts with changes in diffusion tensor imaging (DTI) parameters to assess brain regional susceptibility to diffuse axonal injury (DAI) using thresholds established from in vivo animal and in vitro studies, and to evaluate their performance in assessing the risk of concussion as well as neuropsychological and neurocognitive impairment. These efforts will be facilitated by a large cohort of concussed and matched control athletes with actual on-field head impacts, neuroimaging finding, and clinical symptomatic measures. There are three specific aims. First, we will develop a multi-scale simulation strategy to improve the spatial resolution of the model-estimated responses. In the second aim, we will evaluate the correlation between regional WM fiber strain/strain rate with longitudinal changes in DTI for the targeted athletes based on 1) neuroanatomical regions and 2) prominent WM fiber tracts or neural pathways. In the third aim, we will correlate accumulated WM fiber strain/strain rate in generic and targeted brain regions and neural pathways with neurocognitive alterations as well as clinical diagnosis of concussion. Through these efforts, we will compare the relative performances of WM fiber strain vs. maximum principal strain in generic vs. targeted regions from single vs. repetitive head impacts in injury prediction. We will derive injury thresholds through statistical analysis based on simulation results on concussed and matched control athletes. In addition, we will define, quantify, and rank injury severity index for each prominent WM fiber tract to identify the most vulnerable ones during head impact. This proposal will leverage the established and validated Dartmouth Head Injury Model (DHIM) and an existing large database of actual on-field head kinematics, neuroimaging findings, and clinical outcomes for helmeted athletes generated from previous NIH- and CDC-funded research efforts. The proposed research will accelerate exploration of the biomechanics involved in sports-related concussion on an individual and a group basis which is expected to advance our understanding of which brain regions and/or neural pathways are most susceptible to concussive injury as well as to characterize the head impacts that are more likely to cause such concussive injury and associated neurocognitive impairment.